Genes, emotions and gut microbiota:The next frontier for the gastroenterologist--《World Journal of Gastroenterology》2017年17期
Genes, emotions and gut microbiota:The next frontier for the gastroenterologist
【摘要】：Most medical specialties including the field of gastroenterology are mainly aimed at treating diseases rather than preventing them. Genomic medicine studies the health/disease process based on the interaction of the human genes with the environment. The gastrointestinal(GI) system is an ideal model to analyze the interaction between our genes, emotions and the gut microbiota. Based on the current knowledge, this mini-review aims to provide an integrated synopsis of this interaction to achieve a better understanding of the GI disorders related to bad eating habits and stress-related disease. Since human beings are the result of an evolutionary process, many biological processes such as instincts, emotions and behavior are interconnected to guarantee survival. Nourishment is a physiological need triggered by the instinct of survival to satisfy the body's energy demands. The brain-gut axis comprises a tightly connected neuralneuroendocrine circuitry between the hunger-satiety center, the dopaminergic reward system involved in the pleasure of eating and the gut microbiota that regulates which food we eat and emotions. However, genetic variations and the consumption of high-sugar and high-fat diets have overridden this energy/pleasure neurocircuitry to the point of addiction of several foodstuffs. Consequently, a gut dysbiosis generates inflammation and a negative emotional state may lead to chronic diseases. Balancing this altered processes to regain health may involve personalized-medicine and genome-based strategies. Thus, an integrated approach based on the understanding of the gene-emotions-gut microbiota interaction is the next frontier that awaits the gastroenterologist to prevent and treat GI disorders associated with obesity and negative emotions.
【作者单位】：
【基金】：
【分类号】：R57
欢迎：、、)
支持CAJ、PDF文件格式，仅支持PDF格式
&快捷付款方式
&订购知网充值卡
400-819-9993导读：ExtensivepersonalhumangutmicrobiotaculturecollectionscharacterizedandmanipulatedingnotobioticmiceAndrewL.Goodman1,GeorgeKallstrom,JeremiahJ.Faith,AlejandroReyes,AimeeMoore,GautamDaExtensivepersonalhumangutmicrobiotaculturecollectionscharacterizedandmanipulatedingnotobioticmiceAndrewL.Goodman1,GeorgeKallstrom,JeremiahJ.Faith,AlejandroReyes,AimeeMoore,GautamDantas,andJeffreyI.Gordon2CenterforGenomeScienceandSystemsBiology,WashingtonUniversitySchoolofMedicine,St.Louis,MO63108ContributedbyJeffreyI.Gordon,February24,2011(sentforreviewJanuary21,2011)Theproportionofthehumangutbacterialcommunitythatisrecalcitranttocultureremainspoorlyde?ned.Inthisreport,wecombinehigh-throughputanaerobicculturingtechniqueswithgnotobioticanimalhusbandryandmetagenomicstoshowthatthehumanfecalmicrobiotaconsistslargelyoftaxaandpredictedfunctionsthatarerepresentedinitsreadilyculturedmembers.Whentransplantedintognotobioticmice,completeandculturedcommunitiesexhibitsimilarcolonizationdynamics,biogeograph-icaldistribution,andresponsestodietaryperturbations.More-over,gnotobioticmicecanbeusedtoshapethesepersonalizedculturecollectionstoenrichfortaxasuitedtospeci?cdiets.Wealsodemonstratethatthousandsofisolatesfromasingledonorcanbeclonallyarchivedandtaxonomicallymappedinmultiwellformattocreatepersonalizedmicrobiotacollections.Retrievingcomponentsofamicrobiotathathavecoexistedinsingledonorswhohavephysiologicordiseasephenotypesofinterestandreunitingtheminvariouscombinationsingnotobioticmiceshouldfacilitatepreclinicalstudiesdesignedtodeterminethedegreetowhichtractablebacterialtaxaareabletotransmitdonortraitsorin?uencehostbiology.gutbacterialdiversity|nutrientCmicrobeinteractionstranslationalmedicinepipelineforhumanmicrobiome|Effortstodissectthefunctionalinteractionsbetweenmicrobialcommunitiesandtheirhabitatsarecomplicatedbythelong-standingobservationthat,formanyofthesecommunities,thegreatmajorityoforganismshavenotbeenculturedinthelaboratory(1).Methodologicaldifferencesbetweenculture-independentandculture-basedapproacheshavecontributedtothechallengeofderivingarealisticappreciationofexactlyhowmuchdiscrepancyexistsbetweentheculturablecomponentsofamicrobialecosystemandtotalcommunitydiversity.TableS1givesexamplesofthesemethodologicaldifferences.Thelargestmicrobialcommunityinthehumanbodyresidesinthegut:ItsmicrobiomecontainsatleasttwoordersofmagnitudemoregenesthanarefoundinourHomosapiensgenome(2).Culture-independentmetagenomicstudiesofthehumangutmicrobiotaareidentifyingmicrobialtaxaandgenescorrelatedwithhostphenotypes,butmechanisticandexperimentallydem-onstratedlinksbetweenkeycommunitymembersandspeci?caspectsofhostbiologyaredif?culttoestablishwiththesemethodsalone.Thegoalsofthepresentstudywere(i)toevaluatetherepresentationofreadilyculturedphylotypesint(ii)topro?lethedynamicsoftheseculturedcom-munitiesinamand(iii)todeterminewhetheraclonallyarrayed,personalizedstraincollectioncouldbeconstructedtoserveasafoundationforreassemblingvaryingelementsofahuman’sgutmicrobiotainvitroorinvivo.ResultsToestimatetheabundanceofreadilyculturedbacterialphylo-typesinthedistalhumangut,primerswereusedtoamplify|PNAS|April12,2011|vol.108|no.15variableregion2(V2)ofbacterial16SribosomalRNA(rRNA)genespresentineightfreshlydiscardedfecalsamplesobtainedfromtwohealthy,unrelatedanonymousdonorslivingintheUnitedStates(n=1completesampleperdonoratt=1,2,3,and148d).Ampliconsweresubjectedtomultiplexpyrose-quencing,andtheresultswerecomparedwiththosegeneratedfromDNApreparedfrom～30,000coloniesculturedfromeachsampleunderstrictanaerobicconditionsfor7dat37°Conarichgutmicrobiotamedium(GMM)composedofcommer-ciallyavailableingredients(“cultured”detailsoftheculturingtechniquearegiveninSIMaterialsandMethods,andadescriptionofGMMisgiveninTableS2).Theresulting16SrRNAdatasetswerede-noisedtominimizesequencingerrors(3,4),readsweregroupedintooperationaltaxonomicunits(OTUs)of≥97%nucleotidesequenceidentity(ID),andchimericsequenceswereremoved(SIMaterialsandMethods).Intotal,632distinct97%IDOTUswereobservedinthecom-pletesamples,and316wereidenti?edintheculturedsamples.TheaverageabundanceofculturedOTUsinthecompletesam-pleswas0.4%,buttheaverageabundanceofunculturedOTUs(i.e.,thoseobservedinthecompletebutnottheculturedsamples)wassigni?cantlylower(0.06%;P<10?6byanunpaired,two-tailedStudent’sttest,notassumingequalvariances)(5).Toevaluatetherepresentationofreadilyculturedtaxainthehumangutmicrobiotaatvaryingphylogeneticlevels,weassignedtaxonomicdesignationstoeach97%IDOTU(SIMaterialsandMethodswasscored).Eachas“cultured16SrRNA”ifreadithadfromataxonomicthecompleteassignmentfecalsamplethatalsowasidenti?edinthecorrespondingculturedpopulation.Ifa97%IDOTUinthecompletesamplecouldnotbeplacedinanyknowntaxonomicgroup,itwasscoredas“cultured”onlyifthesame97%IDOTUwasobservedintheculturedsample.Thisanalysisindicatedthat99%ofthe16SrRNAreadsderivedfromthecompletefecalsamplesfromeitherdonorbelongtophylum-,class-andorder-leveltaxathatarealsopresentinthecorresp89±4%ofthereadsarederivedfromreadilyculturedfamily-leveltaxa,and70±5%and56±4?longtoreadilyculturedgenus-andspecies-leveltaxa,re-spectively(Fig.1AUpper).TwoalternatetaxonomicbinningAuthorcontributions:A.L.G.,G.K.,J.J.F.,G.D.,andJ.I.G.A.L.G.,G.K.,J.J.F.,andA.M.A.R.contributednewreagents/A.L.G.,J.J.F.,A.R.,andJ.I.G.andA.L.G.andJ.I.G.wrotethepaper.Theauthorsdeclarenocon?ictofinterest.Datadeposition:ThesequencesreportedinthispaperhavebeendepositedintheNationalCenterforBiotechnologyInformationSequenceReadArchive(accessionnos.SRA270,and026271).FreelyavailableonlinethroughthePNASopenaccessoption.1Presentaddress:SectionofMicrobialPathogenesisandMicrobialDiversityInstitute,YaleUniversity,NewHaven,CT06536.2Towhomcorrespondenceshouldbeaddressed.E-mail:jgordon@wustl.edu.Thisarticlecontainssupportinginformationonlineatwww.pnas.org/lookup/suppl/doi:10.1073/pnas./-/DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.methods,theRibosomalDatabaseProject(RDP)Bayesianclassi?erv2.0andanarbitrary%IDcutoff,producedsimilarresults(Fig.S1ACF).ControlexperimentsdescribedinSIMaterialsgeneratedandfromMethods30,000indicatepooledthatcoloniesatleastare98%notderivedofthereadsfromnongrowingorlysedbacteria(thepercentageofreadsfromtheoriginalfecalsamplesthatarederivedfromdeadcellsisun-known).Unsupervisedhierarchicalclusteringofthecompleteandculturedmicrobialcommunities,acrossthetwodonorsandfourtimepoints,revealedthatculturedsamplesclusterseparatelyfromthosethathadnotbeencultured.Bothphylogeneticandnonphylogeneticmetricssegregateculturedsamplesbydonor,suggestingthatthedistinctivenessofeachdonor’smicrobiotaispreservedintheircollectionsofreadilyculturedrepresentatives(Fig.S1GandH).WeperformedshotgunDNApyrosequencingtodeterminethedegreetowhichpredictedfunctionscontainedinthecom-positegenomesofthecompletehumanfecalmicrobialcom-munitieswererepresentedinthecorrespondingcollectionofculturedmicrobes[n=4samples(onecompleteandonecul-turedfromeachoftwodonors);119,842±43,086high-qualityshotgaveragereadlength,366nt].Onaverage,90%ofthe2,302distinctKEGGOrthology(KO)anno-tationsidenti?edinthetwounculturedsamplesalsowereob-servedintheculturedcommunities(Fig.1B,Fig.S2AandB,andTableS3).Thishighpercentageoffunctionalrepresentationalsowasobservedwhenthemicrobiomesweresubjectedtoalternateannotationschemes.Onaverage,94%of929enzymecommission(EC)assignmentsand95%of216level2KEGGpathwaysasso-ciatedwiththecompletefecalsamplesalsoweredetectedintheculturedcommunities(Fig.S2CCFandTablesS4andS5).Tocomparethefunctionsrepresentedinthecompleteandculturedmicrobiotaindependentofannotation,wecapturedantibiotic-resistancegenesfromtheirmicrobiomesinEscherichiacoli1.5-expressionto4-kBfragmentsvectors.EachofmicrobiomeE.colilibraryDNAcontainedandwas～screened1GBofagainstapanelof15antibioticsandclinicallyrelevantantibioticcombinations(TableS6).Genesencodingresistancetothesame14antibioticswerecapturedinlibrariespreparedfromcompleteandculturedfecalsamples(Fig.S2GandTableS7).Inoneex-ample,ascreenforDNAfragmentsthatconferresistancetotheaminoglycosideamikacinproducedcandidategenesfromthemicrobiomesofbothcompleteandculturedmicrobialcommu-nitiesfromDonor1butnotfromDonor2.Twogenesconferringamikacinresistance(eitherthe16SrRNAmethylasermtDortheaminoglycosidephosphotransferaseaphA-3)wereidenti?edin70%oftheDNAfragmentscapturedinselectionsforthisphe-notype.Directculturingoftheoriginalfecalcommunitiesinthepresenceofamikacincon?rmedthatthisresistancefunctionissigni?cantlyenrichedinthereadilyculturedmicrobiotaofDonor1comparedwithDonor2(P<0.005baseunpaired,two-tailedStudent’sttestassumingequalvariances)(Fig.S2H).PCRanalysisshowedthatmanyoftheamikacin-resistantfecalstrainsharborrmtDoraphA-3.Sequencingthe16SrRNAgenesofasubsetoftheseisolatesindicatedthatrmtDispresentinstrainsofBacteroidesuniformis,B.caccae,andB.thetai-otaomicrontypestrainsinofthisthesedonorspecies)(although,andthatnotably,aphA-3notisincontainedthesequencedinthegenomeofamemberofthegenusDesulfotomaculum(orderClostridiales).Todeterminewhetheracommunitycomposedofanindivid-ual’sreadilyculturedbacteriaexhibitsbehaviorinvivomirroringthatoftheindividual’scompletemicrobialcommunity,9-wk-oldC57Bl6/Jgermfreemicewerecolonizedwithacompleteorculturedmicrobiotafromeachofthetwohumandonors(n=5recipientmicepersampletype).Afecalsamplefromeachdonorwasdividedaftercollection,andonealiquotwasgavageddirectlyGoodmanetal.intoonetheotheraliquotwasculturedonGMMplatesfor7d,asabove,harvested,andintroducedintoasecondgroupofrecipientanimals.Miceweremaintainedonastandardautoclavedlow-fat,plantpolysaccharide-rich(LF/PP)chowdietbeforeand4wkaftergavage.16SrRNAanalysisoffecalsamplescollectedfromthesemiceattheendofthe4-wkperiodindicatedthatthecompleteandtheculturedcommunitieswerein?uencedsimilarlybyhostselection:91±3%ofthe16SrRNAreadsidenti?edfrommicecolonizedwithahumandonor’scompletefecalmicrobiotawerederivedfromgenus-leveltaxathatalsowereidenti?edinthemicecolonizedwiththeculturedmicrobialcommunityfromthesamedonor(Fig.1ALowerharvested,).Importantly,activelygrowingcontrolcoloniesexperimentsgavageddemonstratedintoeachgermfreethatthemouseareabletopreventnongrowingspeciesthatmightbepresentonGMMplatesfromestablishingthemselvesinre-cipientanimals(detailsaregiveninSIMaterialsandMethods).Luminalmaterialwascollectedfromtheproximal,central,anddistalportionsofthesmallintestine,cecum,andcolonofmicecolonizedwitheitherthecompleteorculturedcommunitiesfromeachofthetwohumandonors.V2-directedbacterial16SrRNAsequencingrevealedsimilargeographicvariationsincommunitystructures(Fig.1CandFig.S3ACC).Todeterminewhetherthesimilaritiesincommunitycompo-sitioninvivoextendtosimilaritiesincommunitygenecontent,thesamefecalDNAsamplesthathadbeenpreparedfromthesemiceafter4wkontheLF/PPdietfor16SrRNAanalysesweresubjectedtoshotgunpyrosequencing(n=487,357±30,710readspersample).Aswiththe16SrRNAanalysis,comparisonsoftherepresentationofKOsinthevariousmicro-biomesamplesrevealedanevengreatercorrelationbetweencompleteandculturedcommunitiesaftertheyhadbeensub-jectedtoinvivoselectionthanbeforetheirintroductionintomice(Fig.1B,Fig.S2ACF,andTablesS3CS5).Previouscomparisonsofadultgermfreemicewiththosethatharborgutmicrobialcommunities(eitherconventionallyraisedanimalsorformerlygermfreeanimalscolonizedfrommouseorhumandonors)haveshownthatthepresenceofacompletegutmicrobiotaisassociatedwithincreasedadiposity(6,7).Incomparison,colonizationofgermfreemicewithasingle,readilycultured,prominenthumangutsymbiont(Bacteroidesthetaio-taomicronprominently)orrepresentedwithade?nedincommunitythedistalhumanof12bacterialgut(8)speciesisin-suf?cienttorestoreepididymalfatpadstorestolevelsobservedinconventionallyraisedanimals(datanotshown).Toassesswhetheracomplexcommunityofculturedmicrobescouldrestoreepididymalfatpadweightstothelevelsassociatedwithcompletemicrobialcommunities,weevaluatedmicecolonizedwiththecompleteortheculturedfecalcommunitiesfromthetwohumandonors.Allanimalsdisplayedsigni?cantlygreaterfatpadtobodyweightratiosthangermfreecontrols,andnosigni?cantdifferencewasobservedinadipositybetweenmicecolonizedwiththedonors’completeorculturedmicrobiota(Fig.S3D).Wehavereportedthatmicecolonizedwithacompletehumanmicrobiotaundergomarkedchangesinmicrobialcommunitystructure(evenafterasingleday)whenshiftedfromLF/PPchowtoahigh-fat,high-sugarWesterndiet(7).Totestwhetheramicrobialcommunityconsistingonlyofculturedmembersrecapitulatesthisbehaviorinvivo,thefourgroupsofgnotobioticmicecolonizedwiththecompleteorculturedmicrobesfromtwounrelatedhumandonorsweremonitoredbyfecalsamplingbe-fore,during,andaftera2-wkperiodwhentheywereplacedontheWesterndiet(sampleswerecollectedatdays4,7,and14ofthe?rstLF/PPphase,then1dbeforeand3,7,and14dafterinitiationoftheWesterndietphase,and?nally1,3,8,and15dafterthereturntotheLF/PPdiet).16SrRNA-basedcom-parisonsoffecalcommunitieswereperformedusingbothphy-logeneticandnonphylogeneticdistancemetrics.WitheitherPNAS|April12,2011|vol.108|no.15|6253YGOLOIBORCIMAPercent of complete community0 20 40 60 80 100 phylum class order family In vitrogenus species phylum class order In vivofamily genus species Taxon observed in cultured sample Taxon not observed in cultured sample B250 seitinum200 moc der150 utluc ,ecn100 adnuba 50 OIn vitro, R2 = 0.66KIn vivo, R2 = 0.950 0 50 100 150 200 250 300 KO abundance, complete communitiesC1 V0.8 Verrucomicrobiae etPelp0.6 Deltaproteobacteria emPconC0.4 Betaproteobacteria adn0.2 PuAlphaproteobacteria b0 aF e1 Erysipelotrichi vita0.8 FlClostridia edRer0.6 FutBacilli luC0.4 B0.2 Bacteroidia A0 Actinobacteria (class) Si2Si5Si13CecumColonFecesFig.1.Comparisonofthetaxonomicrepresentationofbacterialspeciesandgenecontentincompleteversusculturedhumanfecalmicrobialcom-munitiesbeforeandaftertheirintroductionintognotobioticmice.(A)16SrRNAsequencesfromcompletemicrobiotawerecomparedwiththoseidenti?edfrommicrobialcommunitiesculturedfromthesamehumandonors.Ateachtaxonomiclevel,theproportionofreadsinthecompletecommunitybelongingtoataxonomicgroupobservedintheculturetheproportionofreadsbelongingtoataxonomicgroupnotobservedintheculturedsample(orlackingtaxonomicassignment)isshowninblack.Datashownaretheaverageoftwounrelatedhumandonors.Invitrosamplesrefertocomparisonsbetweenhumanfecalsamplesandplatedmaterial.Invivosamplesrefertocomparisonsbetweengnoto-bioticmicecolonizedwithacompletehumanfecalmicrobiotaandmicecolonizedwiththereadilyculturedmicrobesfromthesamehumanfecalsample.(B)Annotatedfunctionsidenti?edinthemicrobiomesofcompleteandculturedhumangutcommunities.EachpointrepresentsaKOdesig-nationplottedbyrelativeabundance(averageacrosstwodonors,per100,000sequencingreads).BlackpointsrepresentKOcomparisonsbetworangepointsrepresentcomparisonsbetweeninvivosamples.(C)Thedistributionoftaxaandtheirrelativeabundancealongthelengthoftheintestinearesimilaringnotobioticmicecolonizedwithcom-pleteorculturedhumangutcommunities.Relativeabundancesofclass-level6254|www.pnas.org/cgi/doi/10.1073/pnas.metric,principalcoordinatesanalysis(PCoA)revealedthatmicecolonizedwiththecompleteorculturedsamplesmaintaincommunitiesthatcluster?rstbydonor(principalcoordinate1;PC1)andthatthecompleteandculturedcommunitiesfrombothdonorsrespondtothedietshiftinasimilarmanner[principalcoordinate2(PC2);Fig.2ACCandFig.S4AandB].Likethetransplantedcompletemicrobiotaexaminedhereandinpre-viousreports(7),theculturedmicrobiotarespondedtothisWesterndietbyincreasingtherelativeproportionofrepre-sentativesofoneclassofFirmicutes(theErysipilotrichi)anddecreasingtherelativeproportionoftheBacteroidiaclass(Fig.S4C).Notably,ofthe18species-levelphylotypessigni?cantlyaffectedbydietshiftinthemicecontainingthecompletemicrobiotaofbothhumandonors,14weredetectedanddem-onstratedthesamestatisticallysigni?cantresponseinmicecol-onizedwithreadilyculturedtaxa(Fig.2DandFig.S5).ThefecalmicrobiomesofLF/PP-fedmiceharboringcompleteorculturedcommunitiesfromeachofthetwounrelateddonorswerecomparedwithmicrobiomessampledafterthesemiceconsumedtheWesterndietfor14d(101,222±24,271readspersample).Therepresentationoflevel2KEGGpathwayfunctionswashighlyconcordantonbothdiets,withoneexception:Genesencodingphosphotransferasesystem(PTS)pathwaysforcarbo-hydratetransportweresigni?cantlyoverrepresentedinWesterndiet-fedmiceharboringcompleteorculturedcommunitiesfromeitherdonor(Fig.2EandFig.S6ACD).Higher-resolutionKO-levelannotationscon?rmedthatthediet-basedPTSpathwayenrichmentre?ectedincreasedrepresentationofmultiplecar-bohydratetransporters(Fig.S6EandF).Theseresultsem-phasizethatthesimilartaxonomicrestructuringofcompleteandculturedcommunitiesinresponsetodietisaccompaniedbysimilarchangesincommunitygenecontent.Becausehumangutcommunitiescomposedofreadilycul-turedmembersexhibitresponsestohostdietthatmirrorthosecharacteristicofacompletemicrobiota,weexploredthepossi-bilitythatgnotobioticmicecanbeusedasbiological?lterstorecovercollectionsofreadilyculturedmicrobes,obtainedfromselectedhumanhosts,thatareenrichedforcertainproperties[e.g.,theabilitytoprosper(bloom)whenexposedtospeci?cfoodsorfoodingredients].Tothisend,fecalsamplesfrommicecolonizedwiththecompleteorcorrespondingculturedhumangutmicrobialcommunitiesfromthetwounrelateddonorsandfedtheLF/PPorWesterndietswerecollecteddirectlyintoan-aerobicmediumandthenplatedonprereducedGMMplates(Fig.3A).After7-dincubation,V2-directed16SrRNApro?lingoftheseplatedmicrobialcollectionscon?rmedthatthesepop-ulationsofculturedmicrobescanbereshapeddeliberatelyinvivoandthenrecoveredinvitro(Fig.3BandFig.S7).Oneitherdiet,culturedpopulationsshowedsigni?cantlygreaterresemblancetotheinvivocommunitiesfrommiceconsumingthesamedietthantotheinvivocommunitiesfromthesamemiceconsumingthealter-natediet(P<10?11,unpairedtwo-tailedStudent’sttestofwithin-donordistancesshowninFig.3B,assumingequalvariances).Wenextturnedourattentiontodevelopingwaystodissecttheserecoveredpopulationsthatdisplaypropertiesofinterestinvivo.Thestrictanaerobictechniquesusedhere,compoundedwithhighlydiversecolonymorphologiesacrosstaxa,complicateeffortstopickandisolateindividualcoloniesatascalesuf?cienttocapturethebacterialdiversityrepresentedonthecultureplates.Therefore,weusedamostprobablenumber(MPN)taxaatdatarepresenttheaverageofmicecolonizedfromtwounrelateddonors.Si,smallintestinedividedinto16equal-sizesegmentsandsampledatSi-2(proximal),Si-5(middle),andSi-13(distal).PCoAsuggeststhatgutbiogeography,ratherthandonororculturing,explainsthemajority(58%)ofvariancebetweensamples(Fig.S3ACC).Goodmanetal.ABWestern diet)%9.)7(% 29Donor 2C1( P1CDonor 1PPC3 (6.6)PC1 (19%)DietDonor12InputWesternCulturedCompleteCultured0324662LF/PPTime (Days)CompleteD00.1110100 BacteroidaceaeCWestern dietB.uniformisB.stercorisB.ovatusB.massiliensisB Porphyromonadaceae)%P.goldsteinii9.7LF/PP diet( 2CWesternB PrevotellaceaePF ClostridiaceaeP.claraDQ793820C.DQ803886C.symbiosumC.scindensC.propionicum0324662C.orbiscindensTime (Days)C.leptumC.hathewayiEF Ruminococcaceae6000 R.obeum F.prausnitzii teF Erysipelotrichaceaeid nr3000 nonidentifiedetsPTSF StreptococcaceaeWe2000 F EnterococcaceaeL.lactisPTSE.faeciumP DesulfovibrionaceaeB.wadsworthia1000 V VerrucomicrobiaceaeColonized with complete communityA.muciniphilaColonized with cultured communityTime0 121212121212Donor0 500 00 00 LF/PPWesternLF/PPLF/PPWesternLF/PPDietLF/PP diet Complete microbiotaCultured microbiotaFig.2.Humangutmicrobialcommunitiescomposedonlyofculturedmembersexhibitinvivodynamicssimilartothoseobservedintheircompletecoun-terparts.(A)PCoAofUniFracdistancesbetween16SrRNAdatasetsgeneratedfromfecalsamplesfromgnotobioticmice,colonizedwithcompleteorculturedhumanfecalmicrobialcommunitiesfromtwounrelateddonorsandsampledovertime.Fromday33C46,micewereswitchedfromtheirstandardLF/PPchowtoahigh-fat,high-sugarWesterndiet.Timeseriesanalysisofcommunitystructureasviewedalongthe?rsttwoprincipalcoordinatesfromAshowsthatinterpersonal(donor)differencesseparatedcommunitiesonPC1(B),andhostdietseparatedcommunitiesonPC2(C).Principalcoordinate3(PC3)separatedsamplesfrommicecolonizedwithcompletecommunitiesfromthosecolonizedwithculturedpopulations(Fig.S4A).Nonphylogeneticdistancemetricsproducedsimilarresults(Fig.S4DCH).(D)Evidencethatthecommunityresponsetodietisdrivenbyreadilyculturedbacteriaandthatmembersofthesametaxonomicgroupmanifestdistinctresponsestodietperturbations.Species-leveltaxasigni?cantlyin?uencedbydiet(Student’sttestP≤0.01afterBn=97taxatested)ineitherthecompletecommunities(bluenames),theculturedcommunities(greennames),orboth(rednames)areplottedovertime(arrows).Eachcolumnrepresentstheaveragerelativeabundanceinfecalsamplesharvestedfromthreeto?veindividuallycagedmicethatweresampledatvarioustimes:(i)duringtheinitialLF/PP(ii)duringthesubsequentshifttotheWand(iii)uponreturntoLF/PPchow.Membersoffamily-levelgroupswithatleastonediet-responsivespeciesareshown(excludingrarespecieswithaverageabundance<0.1?rosseachtimepoint).ThenamesofalltaxaareshowninFig.S5.(E)Thefunctionalgenerepertoireinthefecalmicrobiomesofhumanizedgnotobioticmice.EachpointrepresentsaKEGGlevel2thenumberofhitstoeachpathwayper100,000shotgunpyrosequencingreadsisplottedformiceconsumingLF/PP(xaxis)orWestern(yaxis)diets.Datarepresenttheaveragesofmicecolonizedwithmicrobialcommunitiesfromtwounrelateddonors.TheresultsshowthatthefecalmicrobiomeassociatedwiththeWesterndietisenrichedforgenesinpathwaysassociatedwithcarbohydratephosphotransferasesystems(PTS;redarrows)bothinmicecolonizedwithcomplete(uncultured)humangutcommunities(blackpoints)andmicecolonizedwithcommunitiesofreadilyculturedmembers(orangepoints).Donor-speci?cdataandresultsfromalternateannotationschemesareshowninFig.S6andTablesS3CS5.Goodmanetal.PNAS|April12,2011|vol.108|no.15|6255YGOLOIBORCIMAIntroduced intoDonorGF mice directlyor after culturing12LF/PPWLF/PPCultured from mice on LF/PPLF/PPWLF/PPor Western dietsFecal samplescollected from twounrelated donorsBDonor 1Donor 22LF/PP dietCPWestern dietDietDonor12InputWesternCulturedCompleteLF/PPCulturedCompletePC1Fig.3.Thecommunitycompositionofmicrobesculturedfromhumanizedgnotobioticmicecanbereshapedbyalteringhostdiet.(A)Culturecollectionsweregeneratedfromfecalsamplesobtainedfromgnotobioticmicecolonizedwithcompleteorculturedgutmicrobialcommunitiesfromeitheroftwoun-relatedhumandonorsandmaintainedonLF/PPorWesterndiets.(B)PCoAofnonphylogenetic(binaryJaccard)distancesbetweenculturedsamplesindicatesthatmanipulationofhostdietcanbeusedtoshapethecompositionofcom-munitiesrecoveredinculturefromtheseanimals.Analysisofphylogenetic(UniFrac)distancesbetweensamplesproducedsimilarclusteringbydonorandhostdiet(Fig.S7).techniqueforcreatingarrayedspeciescollectionsinamultiwellformatwithoutcolonypicking.We?rstempiricallydeterminedthedilutionpointforafecalsamplethatyields70%emptywells(nodetectablegrowth)afterinoculationinto384-welltraysand7-danaerobicincubation.AssumingthatthedistributionofcellsintothewellsfollowsaPoissondistribution,adilutionthatleaves70%ofwellsemptyshouldyieldnonclonalwells(thatis,wellsthatreceivedmorethanonecellintheinoculum)only5%theremaindershouldbeclonal(Fig.S8FandG).Atthisdilution,ten384-welltraysshouldyield～1,000clonalwells.Wedevelopedthetwo-stepbarcodedpyrosequencingschemeout-linedinFig.4Atoassigna16SrRNAsequencetotheisolate(s)presentineachturbidwell.Weusedthisapproachtocreateanarchived,personalizedculturecollectionoften384-welltraysfromoneofthehumandonors.16SrRNAsequencescouldbeassignedtomorethan99%ofgrowth-positivewells(TableS8).Oneadvantageofclonallyarrayedcollectionsisthattheeffectsof16SrRNAprimerbiasencounteredusingDNAtemplatespreparedfromcomplexmi-crobialpopulationsareminimizedwhenwellscontainasingletaxon.Thispointisillustratedbytheknownbiasofmostcom-monlyusedprimersagainstBi?dobacteriaspp.(9).Membersof6256|www.pnas.org/cgi/doi/10.1073/pnas.thisgenuswerebetterrepresentedamongthesetof16SrRNAgenesproducedfromindividualwellsthanamongthoseobservedincomplexcommunitiesharvestedfromGMMplates.Afterthearchivedtrayshadbeenfrozenunderanaerobicconditionsandstoredat?80°Cfor7mo,recoveryoforganismsfromwellsexceeded60%.Full-length16SrRNAsequencesgeneratedfromtheserecoveredstrainsmatchedtheassignmentsfromthebarcodedpyrosequencingdataineverycase,suggestingthatthedilutionsdidfollowaPoissondistributionaspredicted.Like16SrRNA-basedcommunitypro?ling,suchcollectionsmaymissrare,butimportant,meseedingadditional384-welltrayswiththedilutedsamplewillcaptureadditionalphylotypes(Fig.S8H).Intotal,thisindividual’scul-turecollectioncontained1,172taxonomicallyde?nedisolatesfromfourdifferentphyla,sevenclasses,eightorders,15families,23genera,and48namedbacterialspecies.Novelisolateswereencounteredatthefamily-,genus-,andspecies-levels(TableS8),and69%ofthecompletecommunityhadagenus-levelrepresentativeinthearrayedcollection(Fig.4B).Asaframeofreference,weidenti?edatotalof159humanfecalorgutbac-terialspeciesfromhumansworldwide(includingpathogens)intheGermanResourceCentreforBiologicalMaterial(DSMZ)culturecollection(SIMaterialsandMethods).Assuch,person-alizedmicrobiotacollectionscancomplementthoseofinterna-A1.Anaerobic limiting dilution, 384-well culturing, quantification, and lysis16S rRNA geneV22.PCR #1BC1 (well location within quadrant)Linker sequence3.Pool all samples within each quadrant4.PCR #2BC2 (quadrant and plate)Sequencing adapter5.All sampl combine into a single tube and sequenceB100 n aytmin80 uuh mfom60
tonce 40 lcarecPe20 f 0 phylum class order family genus species Identified in arrayed collectionAbsent in arrayed collectionFig.4.Personalculturecollectionsarchivedinaclonallyarrayed,taxo-nomicallyde?nedformat.(A)Afterlimitingdilutionofthesampleinto384-welltraystothepointatwhichmostturbidwellsareclonal,atwo-step,barcodedpyrosequencingschemeallowseachculturewelltobeassociatedwithitscorrespondingbacterial16SrRNAsequence.Inthe?rstroundofPCR,oneoftheV2-directed16SrRNAprimersincorporates1of96error-correctingbarcodes(BC1,highlightedinred)thatdesignatesthelocation(rowandcolumn)withinaquadrantofthe384-welltraywherethesampleresides.Theprimeralsocontainsa12-bplinker(blue).Allampliconsgen-eratedfromallwellsinagivenquadrantfromasingleplatethenarepooledandsubjectedtoasecondroundofPCRinwhichoneofprimers,whichtargetsthelinkersequence,incorporatesanothererror-correctingbarcode(BC2;green)thatdesignatesthequadrantandplatefromwhichthesampleswerederived,plusanoligonucleotide(gray)usedfor454pyrosequencing.AmpliconsgeneratedfromthesecondroundofPCRthenarepooledfrommultipletraysandsubjectedtomultiplexpyrosequencing.Thisapproachallowsunambiguousassignmentof16SrRNAreadstowellandplateloca-tionsusingaminimumnumbere.g.,96BC1primersand96BC2primersallow962(9,216)wellstobeanalyzed.(B)Representa-tionoftheoriginal(complete)microbialcommunityinthearrayedstraincollection.Goodmanetal.tionalrepositoriesbycapturingstrainsthatcoexistinasharedhabitatwherecommunitystructureandhostparameterscanbemeasured.OurabilitytocapturethislevelofdiversityafterMPNdilutioninthesearrayedcollectionsindicatesthatitisunlikelythatin-terspeciessyntrophicrelationshipsbythemselvesaresuf?cienttoexplainthediversityobservedontheGMMagarplates.Ontheotherhand,thesepersonalizedarrayedculturecollectionsshouldhelpidentifyobligatesyntrophicrelationships(e.g.,byanalyzingthepatternsofco-occurrenceoftaxainwellsharboringmorethanonephylotypeorbycomparingarrayedcollectionsinwhichonesetoftrayscontainsacandidatesyntrophdeliberatelyaddedtoallwells).DiscussionWe?ndthatitispossibletocapturearemarkableproportionofaperson’sfecalmicrobiotausingstraightforwardanaerobiccul-turingconditionsandeasilyobtainedreagents.Variationsinculturingconditions,includingcomponentsthatarenotcom-merciallyavailable(e.g.,sterilerumenorhumanfecalextracts)andotherapproachesformorecloselyapproximatinganativeguthabitat,undoubtedlywillallowadditionalmembersofthehumangutmicrobiotatobeculturedinvitro(10,11).Thesepersonalculturecollectionscanbegeneratedfromhumansrepresentingdiverseculturaltraditionsandvariousphysiologicorpathophysiologicstates.Akeyopportunityisprovidedwhenanaerobiccultureinitiativesarecombinedwithgnotobioticmousemodels,therebyallowingculturecollectionstobechar-acterizedandmanipulatedinmicewithde?ned(includingengineered)genotypeswhoarefeddietscomparabletothoseofthehumandonor,ordietswithsystematicallymanipulatedingredients.Temporalandspatialstudiesofthesecommunitiescanbeusedtoidentifyreadilyculturedmicrobesthatthriveincertainphysiologicalandnutritionalcontexts,creatingadiscoverypipelinefornewprobioticsandforpreclinicalevaluationofthenutritionalvalueoffoodingredients.Basedontheirinvivoresponses,clonallyarchivedculturedrepresentativesofaperson’smicrobiotacanbeselectedforcompletegenomesequencing(includingmultiplestrainsofagivenspecies-levelphylotype)toidentifypotentialfunctionalvariationsthatexistorevolvewithinaspeciesoccupyingagivenhost’sbodyhabitat.CoincidingwiththeintroductionofyetanothergenerationofmassivelyparallelDNAsequencers,thisapproachshouldalsoallowvastscalingofcurrentsequencingeffortsdirectedatcharacterizinghuman(gut)microbialgenomediversity,evolution,andfunction.Inaddition,recoveredorganismscouldalsobeusedassourcematerialforfunctionalmetagenomicscreens(bio-prospecting).Guidedbytheresultsofmetagenomicstudiesofhumanmicrobiotadonors,componentsofapersonalizedcollectionthathavecoevolvedinasinglehostcanbereunitedinvaryingcombinationsingnoto-bioticmice,potentiallyaftergenome-widetransposonmutagen-1.RazumovAS(1932)Mikrobiologija1:131C146.2.QinJ,etal.;MetaHITConsortium(2010)Ahumangutmicrobialgenecatalogueestablishedbymetagenomicsequencing.Nature464:59C65.3.CaporasoJG,etal.(2010)QIIMEallowsanalysisofhigh-throughputcommunitysequencingdata.NatMethods7:335C336.4.ReederJ,KnightR(2010)Rapidlydenoisingpyrosequencingampliconreadsbyexploitingrank-abundancedistributions.NatMethods7:668C669.5.WalkerAW,etal.(2010)Dominantanddiet-responsivegroupsofbacteriawithinthehumancolonicmicrobiota.ISMEJ5:220C230.6.B?ckhedF,etal.(2004)Thegutmicrobiotaasanenvironmentalfactorthatregulatesfatstorage.ProcNatlAcadSciUSA101:.Goodmanetal.esisofselectedtaxaofinterest(8),forfurthermechanisticstudiesoftheirinteractionsandimpactonhostphenotypes.MaterialsandMethodsCulturingofFecalMicrobiota.TheWashingtonUniversityHumanStudiesCommitteereviewedthestudydesign.Freshlydiscardedfecalsamplesfromtwoanonymousunrelatedhumandonorsweretransferredintoananaerobicchamber(CoyLaboratoryProducts)within5minoftheircollectionasde-scribedinSIMaterialsandMethodsandTablesS2andS9.GnotobioticMouseHusbandry.Allexperimentsinvolvingmicewereper-formedwithprotocolsapprovedbytheWashingtonUniversityAnimalStudiesCommittee.GermfreeadultmaleC57BL/6Jmiceweremaintainedinplasticgnotobioticisolators.Colonization,housing,dietmanipulations,andcontrolexperimentstoevaluatethecontributionofunculturedcellstomi-crobialcommunitiesfromgnotobioticmicearedescribedinSIMaterialsandMethods.16SrRNASequencingandAnalysis.TheV2regionofbacterial16SrRNAgeneswassubjectedtoPCRampli?cation(DNAextractionandPCRprotocolsaredescribedinSIMaterialsandMethods).Metadataforall500samples,in-cludingbarcodes,areprovidedinTableS10.All16SrRNApyrosequencingdatasetshavebeendepositedintheNationalCenterforBiotechnologyIn-formation(NCBI)SequenceReadArchive(SRA)(accessionno.SRASrRNAsequenceswere?lteredtoremovelow-qualityorchimericsequences,de-noised,andanalyzedusingQIIMEv1.1(3)withparametersdescribedinSIMaterialsandMethods.Toquantifytherepresentationofculturedandunculturedlineagesinmicrobialcommunities,thepresenceorabsenceofeachphylum-,class-,order-,family-,genus-,orspecies-levelphylotypeassignedtosequencesinthecompletesample(s)wasdeterminedintheculturedsample(s).TaxonomywasassignedwithbothSILVA-VOTE(SIMaterialsandMethodsandTableS11)andRDPBayesianclassi?ers.Datawerenormalizedbytheabundanceofeachtaxonomicgroupintheoriginal(uncultured)sample.Foranalysisofmicrobialcommunitiesfrommice,tax-onomicgroupsobservedinfewerthantworeplicateanimalswereomitted.Protocolsforcreationand16SrRNAsequencingofarrayedculturecollec-tionsaredescribedinSIMaterialsandMethodsandTablesS12andS13.ShotgunPyrosequencing.Aliquots(500ng)ofDNApreparedfromselectedcompleteandculturedmicrobiotawereshearedandligatedtothedefault454Titaniummultiplexidenti?ers(MIDs;RocheRapidLibraryPreparationMethodManual,GSFLXTitaniumSeries,October2009)andsequencedusing454Titaniumpyrosequencingchemistry.After?lteringoflow-qualityorhostDNAsequences,readswerequeriedagainsttheKEGGKOdatabase(v52)usingparametersdescribedinSIMaterialsandMmetadataforeachsampleareprovidedinTableS14,andallshotgunpyrosequencingdatasetsareavailableintheNCBISRAunderaccessionno.SRA026270.Proceduresforbio-prospectingforantibioticresistancegenesincompleteandculturedmicrobialcommunitiesaredescribedinSIMaterialsandMethods.ACKNOWLEDGMENTS.WethankB.Muegge,S.Wagoner,D.O’Donnell,M.Karlsson,M.Gonzalez,A.Keel,T.Ellison,B.Wang,J.Symington,V.Wagner,M.Dunne,andR.Knightforassistanceandcomments.Thisworkwassup-portedbyNationalInstitutesofHealthGrantsDK30292,DK70977,andDK78669(toJ.I.G),F32AI078628andK01DK089121(toA.L.G.),andT32-HD043010(toA.M.)andbytheCrohn’sandColitisFoundationofAmerica.7.TurnbaughPJ,etal.(2009)Theeffectofdietonthehumangutmicrobiome:Ametagenomicanalysisinhumanizedgnotobioticmice.SciTranslMed1:6C14.8.GoodmanAL,etal.(2009)Identifyinggeneticdeterminantsneededtoestablishahumangutsymbiontinitshabitat.CellHostMicrobe6:279C289.9.HillJE,etal.(2010)Improvementoftherepresentationofbi?dobacteriainfecalmicrobiotametagenomiclibrariesbyapplicationofthecpn60universalprimercocktail.ApplEnvironMicrobiol76:.10.NicholsD,etal.(2010)Useofichipforhigh-throughputinsitucultivationof“uncultivable”microbialspecies.ApplEnvironMicrobiol76:.11.BollmannA,LewisK,EpsteinSS(2007)Incubationofenvironmentalsamplesinadiffusionchamberincreasesthediversityofrecoveredisolates.ApplEnvironMicrobiol73:.PNAS|April12,2011|vol.108|no.15|6257YGOLOIBORCIM包含总结汇报、文档下载、资格考试、人文社科、外语学习、行业论文、旅游景点以及Extensive personal human gut microbiota culture collections等内容。本文共2页
相关内容搜索